The new regression equation for the price of NLY is:
Pred. NLY = 12.132 - 1.349(fedfundstargetrate) + 1.238(TYXMA10).
TYXMA10 means that 10-day moving average of the 30-year Treasury bond yield.
The r-sq fell again to 0.41, and the standard error rose from 2.783 to 2.815. The predicted value for NLY (10-day moving average) for 1/5/15 is $15.41. The current price of $11.05 is more than 1.5 standard errors below its mean, and so the stock is cheap.
The sell point is when the price goes beyond 1.5 standard errors above the mean, which is currently at $19.63. This value changes as TYX and FFR change.
A sell also occurs if the yield spread goes above 4.5 percentage points. This is unlikely. The yield spread is currently 2.66, and likely to contract further, because long-term Treasury yields are likely to confound the consensus opinion and fall further as Europe and Japan export their deflation. If the Feds raise short-term rates, that will also contract the yield spread. Of course, raising short-term rates increases the cost of funds for NLY, but since everyone is sure that the Fed Funds rate will rise, they are presumably hedged for it. I think the US inflation rate won't come near the Feds' 2% target, and so short-term rates will not rise.
I have 10% of my portfolio in NLY and 10% in REM. 2013 was bad for the stocks, but 2014 was excellent. NLY's stock price gained 8.4%,and the dividends added $1.20, which comes to another 12% based on the stock price at the end of 2013. I also juiced my returns a little by selling the day before NLY goes ex-dividend, and rebuying when I can get a price less than the price I sold at minus the dividend. For example, I sold on 12/26 at 11.32, eschewed the .30 dividend, and rebought 2 days later at $10.92. So that was equivalent to getting a $.40 dividend. I can do this in tax-free account, and not have to worry about the wash rule.
Historically, you couldn't always get a price less than the closing price prior to the stock going ex-dividend minus the dividend, but usually you could. If the price won't come to you, I suggest that you buy the stock back about 6 trading days before the dividend is paid out. The dividend payout often seems to lift the stock.
This strategy doesn't apply to REM, because it is an open fund, and trades at NAV, which in turn is decided by what the stock prices of the MREITS are, what dividends are due to REM, and what dividends have been paid out by REM. So REM will drop by the dividend amount when it goes ex-dividend, ceteris paribus, but if the component stocks are rising REM may never go back to your target purchase price. REM pays out the dividend about one week after it goes ex-dividend. I re-invest the dividend (sometimes before I get it), if the price drops by more than the dividend payout. If it doesn't, I'll re-invest about the time that NLY and AGNC are paying out their dividends about a month after they go ex-dividend.
REM's stock price grew 1.6% yoy, but the dividends came to 14.85% of the stock price at the end of 2013. Nice.
Disclosure: The author is long NLY, REM.